Nearly every American will experience a diagnostic error in his or her lifetime. Diagnostic errors are particularly damaging when treatable diseases associated with significant morbidity and mortality are missed. Misdiagnosis of cerebrovascular events in the emergency setting can result in lost or delayed opportunities for life-saving treatments. Up to 9% of the estimated 700,000 strokes seen annually in the US are initially misdiagnosed in emergency departments. Among stroke patients who are misdiagnosed, headache is one of the most common presenting symptoms. Though vascular headaches are common among stroke patients, the clear majority of headaches evaluated in US emergency departments are benign. There are few evidence based clinical decision rules to detect acute cerebrovascular disease among patients who complain of headache, particularly for those with non-thunderclap semiology. We therefore propose to quantify the rate of probable misdiagnosis of cerebrovascular disease among headache patients using a large regional information exchange that contains data from multiple institutions based on rates of short-term stroke admission after discharge with a new primary headache diagnosis. We will use information from the Emergency Department of John Hopkins Health System to identify patient and physician-level factors associated with cerebrovascular disease misdiagnosis. Based on prior research, we anticipate that misdiagnosis will be associated with young age and minority race as well as flawed physician clinical reasoning and knowledge gaps in headache medicine. To identify additional sources of diagnostic error, we will conduct a mixed-methods prospective observational study of headache patients evaluated in the Montefiore Medical Center Emergency Department. This work will be augmented by physician interviews at the same center designed to characterize shortcomings in emergency system diagnostic processes. This multi-pronged research approach uses different data sources to address a novel topic that is sure to provide needed insight as to how to improve the diagnosis and treatment of cerebrovascular diseases. We hope this work will pave the way for the development of targeted screening tools, educational interventions, bedside testing, or other low-cost interventions to reduce patient morbidity and mortality by improving diagnostic accuracy. Additionally, our methods of determining rates of cerebrovascular misdiagnosis among headache patients may serve as a model for future symptom-specific studies of diagnostic error in cerebrovascular disease.